// IngestionService: 批次寫入數據以優化資料庫效能
func (s *Server) consumeAndBatchWrite(ctx context.Context) {
    batchSize := 500
    batch := make([]SensorData, 0, batchSize)
    ticker := time.NewTicker(1 * time.Second)

    for {
        select {
        case msg := <-s.kafkaChan:
            batch = append(batch, msg)
            if len(batch) >= batchSize {
                s.dbRepo.BulkInsert(batch) // 批量寫入
                batch = batch[:0] // 清空 Slice
            }
        case <-ticker.C:
            if len(batch) > 0 {
                s.dbRepo.BulkInsert(batch) // 定時強制寫入，避免數據滯留
                batch = batch[:0]
            }
        case <-ctx.Done():
            return
        }
    }
}

import pandas as pd
import pandas_ta as ta  # 引入技術分析庫

class StrategyEngine:
    def __init__(self, df: pd.DataFrame):
        self.df = df

    def strategy_golden_cross(self):
        """
        策略：KD 黃金交叉 + 成交量放大
        """
        # 1. 計算 K, D 值 (9, 3, 3)
        kds = ta.stoch(self.df['high'], self.df['low'], self.df['close'], k=9, d=3)
        self.df = pd.concat([self.df, kds], axis=1)
        
        # 2. 定義訊號邏輯
        # 當天 K > D 且 昨天 K < D (黃金交叉)
        signal_cross = (self.df['STOCHk_9_3_3'] > self.df['STOCHd_9_3_3']) & \
                       (self.df['STOCHk_9_3_3'].shift(1) < self.df['STOCHd_9_3_3'].shift(1))
        
        # 當天成交量 > 5日均量 * 1.5 (爆量)
        vol_ma5 = self.df['volume'].rolling(window=5).mean()
        signal_volume = self.df['volume'] > (vol_ma5 * 1.5)

        # 3. 篩選出符合條件的日期
        signals = self.df[signal_cross & signal_volume]
        
        return signals

# 模擬運行
# data = get_stock_data("2330.TW")
# engine = StrategyEngine(data)
# buy_signals = engine.strategy_golden_cross()
# if not buy_signals.empty:
#     send_line_notify(f"發現買入訊號: 台積電 at {buy_signals.index[-1]}")

pip install yfinance mplfinance pandas

import yfinance as yf
import mplfinance as mpf
import pandas as pd

def fetch_and_plot_stock(stock_code, period="1y", interval="1d"):
    """
    功能：抓取台股數據並繪製專業 K 線圖
    參數：
        stock_code (str): 股票代碼 (例如 "2330")
        period (str): 期間 (例如 "1y", "6mo", "3mo")
        interval (str): 頻率 (例如 "1d", "1wk")
    """
    
    # 1. 格式化股票代碼 (台股需加上 .TW)
    target_stock = stock_code.upper()
    if not target_stock.endswith(".TW") and not target_stock.endswith(".TWO"):
        # 簡單判斷：預設嘗試上市 (.TW)，若查無資料可改為上櫃 (.TWO)
        target_stock = f"{stock_code}.TW"
    
    print(f"🚀 Devin 正在從 Yahoo Finance 抓取 {target_stock} 的數據...")

    # 2. 抓取數據
    try:
        df = yf.download(target_stock, period=period, interval=interval, progress=False)
        
        # 檢查數據是否為空
        if df.empty:
            print(f"❌ 錯誤：找不到股票 {target_stock} 的數據。請確認代碼是否正確 (上櫃股票試著手動加 .TWO)。")
            return
        
        # 數據清洗：確保索引是 Datetime 格式
        df.index = pd.to_datetime(df.index)
        
        print(f"✅ 數據獲取成功！共 {len(df)} 筆交易日資料。")
        print(f"   最新收盤價 ({df.index[-1].date()}): {df['Close'].iloc[-1]:.2f}")

    except Exception as e:
        print(f"❌ 發生錯誤: {e}")
        return

    # 3. 設定繪圖風格與均線
    # style='yahoo' 為傳統紅綠K線 (但在國際版 Yahoo 是 綠漲/紅跌，我們稍後自訂顏色)
    # mav=(5, 20, 60) 自動計算並繪製 5日, 20日, 60日 移動平均線
    
    # 自定義台股風格 (紅漲綠跌)
    mc = mpf.make_marketcolors(up='r', down='g', inherit=True)
    s  = mpf.make_mpf_style(base_mpf_style='yahoo', marketcolors=mc)

    # 4. 繪製圖表
    print("📊 正在生成 K 線圖...")
    mpf.plot(
        df,
        type='candle',       # 繪製蠟燭圖
        style=s,             # 套用台股風格
        mav=(5, 20, 60),     # 繪製均線
        volume=True,         # 顯示成交量
        title=f"Stock Analysis: {target_stock}",
        ylabel="Price (TWD)",
        ylabel_lower="Volume",
        figratio=(16, 9),    # 寬螢幕比例
        figscale=1.2,        # 放大圖表
        tight_layout=True
        # savefig='stock_chart.png' # 如果想存成圖片，解開這行註解
    )
    print("✨ 圖表繪製完成！")

# --- 執行區 ---
if __name__ == "__main__":
    # 你可以在這裡修改想要查詢的股票代碼
    # 比如：2330 (台積電), 3081 (聯亞), 8401 (白紗科)
    stock_id = input("請輸入股票代碼 (例如 2330): ").strip()
    if stock_id:
        fetch_and_plot_stock(stock_id)
    else:
        fetch_and_plot_stock("2330") # 預設台積電

pip install flet yfinance mplfinance matplotlib

import flet as ft
import yfinance as yf
import mplfinance as mpf
import pandas as pd
import io
import base64
import matplotlib
matplotlib.use('Agg')  # 設定後端，避免彈出視窗，只生成圖片數據

def main(page: ft.Page):
    # --- 1. APP 基礎設定 ---
    page.title = "AlphaPulse Mobile"
    page.theme_mode = ft.ThemeMode.DARK  # 金融軟體標配：深色模式
    page.padding = 20
    page.vertical_alignment = ft.MainAxisAlignment.START
    
    # 設置 APP 視窗大小 (模擬手機)
    page.window_width = 400
    page.window_height = 800

    # --- 2. 邏輯處理函數 ---
    
    def get_stock_data(e):
        stock_code = txt_stock_id.value.strip()
        if not stock_code:
            return

        # 顯示讀取條
        progress_ring.visible = True
        btn_analyze.disabled = True
        chart_image.src_base64 = None
        price_text.value = "查詢中..."
        change_text.value = ""
        page.update()

        try:
            # 格式化代碼 (自動加 .TW)
            target = stock_code.upper()
            if not target.endswith(".TW") and not target.endswith(".TWO"):
                target = f"{stock_code}.TW"

            # 抓取數據
            df = yf.download(target, period="6mo", interval="1d", progress=False)
            
            if df.empty:
                price_text.value = "查無資料"
                change_text.value = "請確認代碼"
                change_text.color = ft.colors.GREY
            else:
                # --- 更新價格資訊 ---
                latest_price = df['Close'].iloc[-1]
                prev_price = df['Close'].iloc[-2]
                diff = latest_price - prev_price
                percent = (diff / prev_price) * 100
                
                price_text.value = f"${latest_price:.2f}"
                change_text.value = f"{diff:+.2f} ({percent:+.2f}%)"
                
                # 根據漲跌變色 (台股：紅漲綠跌)
                if diff > 0:
                    change_text.color = ft.colors.RED_400
                    price_text.color = ft.colors.RED_400
                elif diff < 0:
                    change_text.color = ft.colors.GREEN_400
                    price_text.color = ft.colors.GREEN_400
                else:
                    change_text.color = ft.colors.WHITE
                    price_text.color = ft.colors.WHITE

                # --- 繪製 K 線圖 ---
                # 設定圖表風格
                mc = mpf.make_marketcolors(up='r', down='g', inherit=True)
                s  = mpf.make_mpf_style(base_mpf_style='nightclouds', marketcolors=mc) # 使用深色風格

                # 將圖表存入記憶體 Buffer
                buf = io.BytesIO()
                mpf.plot(df, type='candle', style=s, mav=(5, 20), volume=True, 
                         title=f'\n{target}', savefig=dict(fname=buf, dpi=100, bbox_inches='tight'))
                
                # 轉為 Base64 字串供 Flet 顯示
                buf.seek(0)
                b64_string = base64.b64encode(buf.read()).decode()
                chart_image.src_base64 = b64_string
                chart_image.visible = True

        except Exception as ex:
            price_text.value = "錯誤"
            change_text.value = str(ex)
        
        # 恢復 UI 狀態
        progress_ring.visible = False
        btn_analyze.disabled = False
        page.update()

    # --- 3. UI 元件佈局 ---
    
    # 標題
    title_row = ft.Row(
        [ft.Icon(ft.icons.CANDLESTICK_CHART, color=ft.colors.BLUE_400),
         ft.Text("AlphaPulse", size=24, weight=ft.FontWeight.BOLD)],
        alignment=ft.MainAxisAlignment.CENTER
    )

    # 輸入區
    txt_stock_id = ft.TextField(
        label="輸入股票代碼 (例如 2330)", 
        border_radius=10,
        text_align=ft.TextAlign.CENTER,
        width=200,
        on_submit=get_stock_data # 按 Enter 觸發
    )
    
    btn_analyze = ft.ElevatedButton(
        text="分析", 
        icon=ft.icons.SEARCH, 
        on_click=get_stock_data,
        bgcolor=ft.colors.BLUE_600,
        color=ft.colors.WHITE
    )

    # 價格顯示區
    price_text = ft.Text("-", size=40, weight=ft.FontWeight.BOLD)
    change_text = ft.Text("", size=18)
    price_container = ft.Column(
        [price_text, change_text], 
        horizontal_alignment=ft.CrossAxisAlignment.CENTER
    )

    # 圖表區
    chart_image = ft.Image(
        src_base64=None, 
        width=380, 
        height=300, 
        fit=ft.ImageFit.CONTAIN,
        visible=False,
        border_radius=10
    )
    
    progress_ring = ft.ProgressRing(visible=False)

    # --- 4. 組裝頁面 ---
    page.add(
        ft.Container(height=20), # 間距
        title_row,
        ft.Divider(color=ft.colors.GREY_800),
        ft.Row([txt_stock_id], alignment=ft.MainAxisAlignment.CENTER),
        ft.Row([btn_analyze], alignment=ft.MainAxisAlignment.CENTER),
        ft.Container(height=20),
        progress_ring,
        price_container,
        ft.Container(height=10),
        chart_image,
    )

# 啟動 APP
ft.app(target=main)

pip install flet yfinance mplfinance pandas matplotlib

import flet as ft
import yfinance as yf
import mplfinance as mpf
import pandas as pd
import sqlite3
import io
import base64
import matplotlib

# 設定 Matplotlib 後端，避免彈出視窗
matplotlib.use('Agg')

# --- 資料庫層 (Model Layer) ---
class Database:
    def __init__(self):
        self.conn = sqlite3.connect("alphapulse.db", check_same_thread=False)
        self.cursor = self.conn.cursor()
        self.create_table()

    def create_table(self):
        self.cursor.execute("""
            CREATE TABLE IF NOT EXISTS watchlist (
                symbol TEXT PRIMARY KEY,
                name TEXT
            )
        """)
        self.conn.commit()

    def add_stock(self, symbol):
        try:
            self.cursor.execute("INSERT INTO watchlist (symbol) VALUES (?)", (symbol,))
            self.conn.commit()
            return True
        except sqlite3.IntegrityError:
            return False # 已存在

    def remove_stock(self, symbol):
        self.cursor.execute("DELETE FROM watchlist WHERE symbol = ?", (symbol,))
        self.conn.commit()

    def get_watchlist(self):
        self.cursor.execute("SELECT symbol FROM watchlist")
        return [row[0] for row in self.cursor.fetchall()]

db = Database()

# --- 業務邏輯層 (Service Layer) ---
def fetch_stock_info(symbol):
    """獲取單檔股票即時價格與變動"""
    try:
        ticker = yf.Ticker(symbol)
        # 獲取最後兩天的數據來計算漲跌
        df = ticker.history(period="5d") 
        if df.empty or len(df) < 2:
            return None
        
        current = df['Close'].iloc[-1]
        prev = df['Close'].iloc[-2]
        change = current - prev
        pct = (change / prev) * 100
        
        return {
            "symbol": symbol,
            "price": current,
            "change": change,
            "pct": pct
        }
    except:
        return None

def generate_chart(symbol):
    """生成 K 線圖 Base64"""
    try:
        df = yf.download(symbol, period="6mo", interval="1d", progress=False)
        if df.empty: return None
        
        mc = mpf.make_marketcolors(up='r', down='g', inherit=True)
        s = mpf.make_mpf_style(base_mpf_style='nightclouds', marketcolors=mc)
        
        buf = io.BytesIO()
        mpf.plot(df, type='candle', style=s, mav=(5, 20, 60), volume=True,
                 title=f'\n{symbol}', savefig=dict(fname=buf, dpi=100, bbox_inches='tight'))
        buf.seek(0)
        return base64.b64encode(buf.read()).decode()
    except:
        return None

# --- UI 層 (View Layer) ---
def main(page: ft.Page):
    # APP 設定
    page.title = "AlphaPulse Pro"
    page.theme_mode = ft.ThemeMode.DARK
    page.padding = 0 # 全螢幕感
    page.window_width = 400
    page.window_height = 850
    
    # 顏色定義
    COLOR_UP = ft.colors.RED_400
    COLOR_DOWN = ft.colors.GREEN_400
    COLOR_BG = ft.colors.GREY_900
    COLOR_CARD = ft.colors.GREY_800

    # --- 元件宣告 ---
    
    # 1. 自選股列表容器
    lv_watchlist = ft.ListView(expand=True, spacing=10, padding=20)

    # 2. 分析頁面元件
    txt_search = ft.TextField(
        hint_text="輸入代碼 (如 2330)", 
        border_radius=30, 
        content_padding=20, 
        expand=True
    )
    img_chart = ft.Image(src_base64=None, fit=ft.ImageFit.CONTAIN, visible=False, animate_opacity=300)
    lbl_price = ft.Text("-", size=32, weight="bold")
    lbl_change = ft.Text("-", size=16)
    btn_add_fav = ft.IconButton(icon=ft.icons.STAR_BORDER, icon_size=30, disabled=True)
    loading_ring = ft.ProgressRing(visible=False)
    
    container_analysis = ft.Column(
        scroll=ft.ScrollMode.AUTO,
        controls=[
            ft.Container(height=20),
            ft.Row([txt_search, ft.IconButton(icon=ft.icons.SEARCH, on_click=lambda e: analyze_stock(txt_search.value))], alignment="center"),
            ft.Container(height=20),
            loading_ring,
            ft.Column([lbl_price, lbl_change], horizontal_alignment="center"),
            ft.Container(height=10),
            ft.Row([btn_add_fav], alignment="center"), # 加入自選按鈕
            ft.Container(height=10),
            img_chart
        ],
        horizontal_alignment="center",
        padding=20,
        visible=False # 預設隱藏，先顯示自選股
    )

    container_watchlist = ft.Column(
        controls=[
            ft.Container(
                content=ft.Text("我的自選股", size=28, weight="bold"),
                padding=ft.padding.only(left=20, top=40, bottom=10)
            ),
            lv_watchlist
        ],
        expand=True,
        visible=True
    )

    # --- 功能函數 ---

    def format_symbol(code):
        code = code.strip().upper()
        if code.isdigit(): return f"{code}.TW"
        return code

    def build_stock_card(data):
        """建立單一股票卡片"""
        color = COLOR_UP if data['change'] > 0 else COLOR_DOWN
        icon = ft.icons.TRENDING_UP if data['change'] > 0 else ft.icons.TRENDING_DOWN
        
        return ft.Dismissible(
            key=data['symbol'],
            on_dismiss=lambda e: delete_stock(e, data['symbol']),
            background=ft.Container(bgcolor=ft.colors.RED, content=ft.Icon(ft.icons.DELETE), alignment=ft.alignment.center_right, padding=20),
            content=ft.Card(
                color=COLOR_CARD,
                elevation=5,
                content=ft.Container(
                    padding=15,
                    content=ft.Row([
                        ft.Column([
                            ft.Text(data['symbol'], size=18, weight="bold"),
                            ft.Text("TW Stock", size=12, color=ft.colors.GREY_500),
                        ]),
                        ft.Row([
                            ft.Column([
                                ft.Text(f"${data['price']:.2f}", size=18, weight="bold", text_align="right"),
                                ft.Text(f"{data['change']:+.2f} ({data['pct']:+.2f}%)", color=color, size=14, text_align="right"),
                            ], alignment="end"),
                            ft.Icon(icon, color=color)
                        ])
                    ], alignment="spaceBetween")
                )
            )
        )

    def load_watchlist():
        """讀取並刷新自選股列表"""
        lv_watchlist.controls.clear()
        loading_ring.visible = True # 借用一下
        page.update()
        
        stocks = db.get_watchlist()
        if not stocks:
            lv_watchlist.controls.append(ft.Text("尚未加入自選股，請去搜尋頁新增！", color=ft.colors.GREY_500, text_align="center"))
        else:
            for symbol in stocks:
                data = fetch_stock_info(symbol)
                if data:
                    lv_watchlist.controls.append(build_stock_card(data))
        
        loading_ring.visible = False
        page.update()

    def delete_stock(e, symbol):
        db.remove_stock(symbol)
        page.show_snack_bar(ft.SnackBar(content=ft.Text(f"{symbol} 已移除"), action="復原"))
        # 注意：這裡簡單處理，實際上應該要有復原邏輯，這裡僅移除
        
    def add_to_watchlist(e):
        symbol = format_symbol(txt_search.value)
        if db.add_stock(symbol):
            btn_add_fav.icon = ft.icons.STAR
            btn_add_fav.icon_color = ft.colors.YELLOW
            page.show_snack_bar(ft.SnackBar(content=ft.Text(f"已將 {symbol} 加入自選清單！")))
            page.update()
        else:
            page.show_snack_bar(ft.SnackBar(content=ft.Text(f"{symbol} 已經在清單中了")))

    def analyze_stock(code):
        if not code: return
        symbol = format_symbol(code)
        
        # UI 狀態重置
        loading_ring.visible = True
        img_chart.visible = False
        lbl_price.value = "分析中..."
        lbl_change.value = ""
        btn_add_fav.disabled = True
        btn_add_fav.icon = ft.icons.STAR_BORDER
        btn_add_fav.icon_color = ft.colors.WHITE
        page.update()

        # 數據獲取
        data = fetch_stock_info(symbol)
        b64_chart = generate_chart(symbol)

        if data and b64_chart:
            lbl_price.value = f"${data['price']:.2f}"
            color = COLOR_UP if data['change'] > 0 else COLOR_DOWN
            lbl_change.value = f"{data['change']:+.2f} ({data['pct']:+.2f}%)"
            lbl_change.color = color
            lbl_price.color = color
            
            img_chart.src_base64 = b64_chart
            img_chart.visible = True
            
            # 啟用加入最愛按鈕
            btn_add_fav.disabled = False
            btn_add_fav.on_click = add_to_watchlist
            
            # 檢查是否已在清單中
            if symbol in db.get_watchlist():
                btn_add_fav.icon = ft.icons.STAR
                btn_add_fav.icon_color = ft.colors.YELLOW
        else:
            lbl_price.value = "查無資料"
            lbl_change.value = "請確認代碼"
        
        loading_ring.visible = False
        page.update()

    # --- 導航邏輯 ---
    def on_nav_change(e):
        index = e.control.selected_index
        if index == 0:
            container_watchlist.visible = True
            container_analysis.visible = False
            load_watchlist() # 切換回首頁時刷新
        else:
            container_watchlist.visible = False
            container_analysis.visible = True
        page.update()

    # --- 底部導航欄 ---
    page.navigation_bar = ft.NavigationBar(
        destinations=[
            ft.NavigationDestination(icon=ft.icons.DASHBOARD_OUTLINED, selected_icon=ft.icons.DASHBOARD, label="自選監控"),
            ft.NavigationDestination(icon=ft.icons.ANALYTICS_OUTLINED, selected_icon=ft.icons.ANALYTICS, label="個股分析"),
        ],
        on_change=on_nav_change
    )

    # --- 啟動初始化 ---
    page.add(
        ft.Stack([
            container_watchlist,
            container_analysis
        ])
    )
    load_watchlist()

ft.app(target=main)

你的專案/
├── alphapulse_ultimate.py
└── .github/
    └── workflows/
        └── build.yml

#### 4. 下載你的軟體
1.  上傳完成後，點擊 GitHub 頁面上方的 **"Actions"** 標籤。
2.  你會看到一個叫做 "Build AlphaPulse EXE" 的任務正在旋轉（執行中）。
3.  等待約 3-5 分鐘，變成綠色勾勾後，點進去。
4.  滑到最下方 **"Artifacts"** 區塊，你會看到 `AlphaPulse-Windows-App`。
5.  點擊下載，解壓縮後，裡面就是不需要 Python 也能執行的 `AlphaPulse.exe` 了！

---

### 如果你堅持要在「這台」電腦上打包...

如果不想用 GitHub，你就**必須**安裝 Python。這是無法繞過的步驟。

1.  去 [Python 官網下載](https://www.python.org/downloads/) Windows Installer。
2.  **重要**：安裝時務必勾選 **"Add Python to PATH"**。
3.  安裝完後，打開 CMD (命令提示字元)，輸入：
    ```bash
    pip install flet yfinance mplfinance pandas matplotlib pyinstaller
    flet pack alphapulse_ultimate.py --name AlphaPulse --onefile
    
**建議選擇第一種「GitHub Actions」方法，這是最現代化且乾淨的開發流程。**